- Email: [email protected]
Homogeneous catalysis of soybean oil transesterification via methylic and ethylic routes: Multivariate comparison
Energy 67 (2014) 569e574
Contents lists available at ScienceDirect
Energy journal homepage: www.elsevier.com/locate/energy
Homogeneous catalysis of soybean oil transesterification via methylic and ethylic routes: Multivariate comparison Karen Araújo Borges a, André Luiz Squissato a, Douglas Queiroz Santos a, Waldomiro Borges Neto a, Antônio Carlos Ferreira Batista b, Tiago Almeida Silva c, Andressa Tironi Vieira d, Marcelo Firmino de Oliveira d, *, Manuel Gonzalo Hernández-Terrones a a
Instituto de Química, Universidade Federal de Uberlândia, Campus Santa Mônica, 38400-902, Uberlândia, Minas Gerais, Brazil Laboratório de Energias Renováveis e Meio ambiente do Pontal (LERMAP), Universidade Federal de Uberlândia, 38302-000, Ituiutaba, MG, Brazil c Departamento de Química, Universidade Federal de São Carlos, 13560-970, São Carlos, SP, Brazil d Departamento de Química, Faculdade de Filosofia, Ciências e Letras de Ribeirão Preto, Universidade de São Paulo, 14040-901, Ribeirão Preto, SP, Brazil b
a r t i c l e i n f o
a b s t r a c t
Article history: Received 21 September 2013 Received in revised form 31 January 2014 Accepted 2 February 2014 Available online 28 February 2014
An experiment to establish the best reaction conditions for the transesterification of soybean oil is described. We conducted the ethylic and methylic routes using two different protocols, and evaluated how the variables time, stirring, alcohol/oil molar ratio, catalyst (%), catalyst type, and temperature affected the process. The highest yield of biodiesel was obtained using the following conditions: ethylic route e t ¼ 60 min, stirring: 100 rpm, ethanol/oil molar ratio ¼ 12:1, catalyst relative to oil (%) ¼ 0.2%, catalyst ¼ potassium ethoxide, temperature ¼ 35 C; methylic route e t ¼ 30 min, stirring: 100 rpm, methanol/oil molar ratio ¼ 6:1, catalyst (%) ¼ 0.2%, catalyst ¼ KOH, temperature ¼ 55 C. We analyzed the acidity, moisture content, density at 20 C, kinematic viscosity at 40 C, oxidative stability, and carbon residue at the biodiesels obtained under optimal conditions. The results were consistent with the values allowed by the Brazilian ANP (Resolution 07/2008). We also conducted the physicochemical analysis of the soybean oil used as feedstock to produce biodiesel. Ó 2014 Elsevier Ltd. All rights reserved.
Keywords: Transesterification Soybean oil Optimization Biodiesel
1. Introduction The oil crisis seen in recent decades together with growing demand for fuel and increasing concern about the environment has prompted the search for alternative energy sources both in Brazil and worldwide [1,2]. Research has focused on developing new basic inputs of renewable character, to produce fuels that can replace petroleum products. In this context, biomass plays a leading role. It is renewable, widely available, biodegradable, and inexpensive [1,2]. Indeed, plant oils had already been tested and used as fuel in diesel engines before the advent of petroleum diesel. However, for both economic and technical reasons, they gave way to diesel oil [3,4]. Currently, biodiesel is a well established example of the use of biomass to produce energy and offers advantages over petroleum diesel [5e7]. It is non-toxic, originates from renewable sources, and
* Corresponding author. Tel./fax: þ55 34 3239 4425. E-mail addresses: [email protected], [email protected], fl[email protected] (M.F. de Oliveira). http://dx.doi.org/10.1016/j.energy.2014.02.012 0360-5442/Ó 2014 Elsevier Ltd. All rights reserved.
leads to better quality of emissions during the combustion process [8]. Although biodiesel provides about 10% less energy than diesel fuel, motor performance is essentially the same with respect to power and torque [6]. Additionally, biodiesel is highly viscous, enhancing lubricity and reducing wear of the moving parts of the engine. Several processes exist to produce biodiesel, but transesterification is the method that is often employed most worldwide [9e11]. It involves reacting a lipid (known as triglycerides or triacylglycerols) with a mono-short chain alcohol (methyl or ethyl) in the presence of a catalyst (acid or base), which produces a mixture of alkyl esters of fatty acids (known as biodiesel) and glycerol [9e11]. The transesterification reaction has been conduced employing different catalytic routes. The catalysts used for the transesterification can be classified into: homogeneous and heterogeneous (basic and acid), or biological (enzymes) [9,12,13]. Among all catalytic routes, the basic homogeneous catalysts, typically sodium and potassium hydroxide, in fact have been most often used industrially, by various as reasons, such as [9,12e14]: low cost catalyst, high catalytic activity with maximum conversion achieved
570
K.A. Borges et al. / Energy 67 (2014) 569e574
in a minimal time, simple operational due to the mild conditions, and minor problem with the corrosive effects. Many factors affect the transesterification of oils and fats; for example, the presence of free fatty acids, humidity, alcohol type, molar alcohol/oil ratio, catalyst type and concentration, reaction time, and temperature [15,16]. Therefore, studies involving transesterification of vegetable oil must be performed considering the effect of the various factors during the optimization, and simultaneously, because the experimental parameters are interrelated. This necessity/approach can be achieved from the use factorial design, where from a small number of experiments diverse variables are studies simultaneously at different levels and taking into account its interactions with the others variables [17,18]. Thus, the recent literature reports the application of factorial design for the optimization of the homogeneous catalysis transesterification from different feedstocks, such as lard [17], castor oil [19], sunflower oil [20], beef tallow [21], Muskmelon (Cucumis melo) seed oil [22], waste vegetable oil [23], among others. In regard the biodiesel production from soybean oil the traditional univariate methodology for optimization of the process has been adopted [24]. However, Oliveira et al. [25] presented the optimization of alkaline transesterification of soybean oil using a Taguchi experimental design, and from this experimental design the effect of the variables temperature, reaction time, catalyst concentration and oil-toethanol molar ratio were available. The influence of other very important factor were not considered in this reported work, as type of alcohol and basic catalyst [25]. In this work a novel approach for the homogeneous catalysis optimization of soybean oil transesterification via methylic and ethylic routes is presented, which all experimental parameters were systematically evaluated using a factorial design. Thus we investigated which of the variables reaction time, reaction rate, molar alcohol/oil ratio, catalyst type, catalyst concentration, and temperature significantly influence the yield of the transesterification reaction using a fractional factorial design 262 in duplicate, which resulted in 32 experiments conducted for each design. 2. Experimental 2.1. Characterization of the soybean oil and biodiesel Soybean oil was analyzed for the following parameters: refractive index (40 C), saponification index, acidity, moisture content, density at 20.0 C, kinematic viscosity at 40 C; oxidative stability, and carbon residue. The same analyses were also conducted for the biodiesel obtained in the optimum experimental conditions, except for the refractive index and saponification number. The parameters refractive index (40 C), saponification number, and acidity were determined according to the official procedures recommended by the American Oil Chemists Society [26]. The moisture content was analyzed according to the norm ASTM D-6304 using a Karl Fischer colorimetric titrator model 831 KF. The density was determined according to the norm ASTM D-4052, which corresponds to the Brazilian norm ABNT NBR 14065, using a DA-500-Kyoto densimeter. Kinematic viscosity at 40 C was obtained according to the norms ASTM D-445 and ASTM 446. The oxidative stability was analyzed by the method EN 14112, on a Rancimat equipment model 743 from Metrohm. 2.2. Transesterification of the soybean oil Refined commercial soybean oil, household, packaging 900 mL, model Liza, industrially processed by Cargill Agricultural SA was used. Table 1 shows the fatty acid composition of the oil.
For the transesterification reaction, the catalyst was added to the solvent; the mixture was stirred until complete dissolution and transferred to a 250 mL Erlenmeyer flask. Next, the oil was added, and transesterification was performed by following the experimental conditions described in the experimental design ethyl 1 and 2 and methyl 1 and 2 (Table 2). After transesterification, two phases were noted: upper and lower crude biodiesel glycerin. Then, biodiesel was removed from the mixture and washed five times with water at 80 C at a water/biodiesel ratio 1:3 (v/v), to remove impurities. The biodiesel was then dried in a rotary evaporator. The wash water containing the residue catalyst was properly neutralized with 0.1 mol L1 HCl solution before being discarded. 2.3. Experimental design To establish the experimental conditions that improved the efficiency of transesterification, the reaction was carried out using a fractional factorial experimental design, in duplicate. The option used in the experiment was 262, which is interesting for investigative purposes, because it reduces the number of tests. This design is not completely saturated and no main effects mix with first order, which ensures that the effects of the variables analyzed in the response are reliably calculated, without statistical information quality. Table 2 lists the values employed at each level of the variables, chosen on the basis of literature studies [27,28]. 3. Results and discussion 3.1. Experimental design 3.1.1. Optimization of the ethylic route using KOH or NaOH as catalyst Tables 3e6 present the corresponding values of mass production regarding the methylic and ethylic transesterification of soybean oil conducted according to the experimental design 262. We compared the performance of the catalysts KOH, NaOH, and potassium ethoxide in the ethylic route of soybean oil transesterification. We also compared the use of KOH, NaOH, and potassium methoxide as catalysts in the methylic route. We calculated the mass yield of the process on the basis of the transesterification reaction stoichiometry: 1 mol of oil affords 3 mol of ester (Table 7). Fig. 1 represents the Pareto chart obtained from the data summarized in Tables 3e6 using the fractional factorial design; we evaluated the effects of each variable and their interactions on the yield of soybean oil transesterification. The Pareto chart corresponding to the design shown in Table 3 (Fig. 1a) revealed that the interactions between factors (1) and (4), (1) and (3), (2) and (5), (1) and (5) affected transesterification less than the variables disregarded then in subsequent statistical analyses.
Table 1 Composition of fatty acids in soybean. Fatty acids
Structure
Reference values (%)
e Myristic Palmitic Palmitoleic Estearic Oleic (Omega 9) Linoleic (Omega 6) Linolenic (Omega 3) Araquidic Eicosenoic Behenic
C<14 C14:0 C16:0 C16:1 C18:0 C18:1 C18:2 C18:3 C20:0 C20:1 C22:0
<0.1 <0.5 7.0e14.0 <0.5 1.4e5.5 19.0e30.0 44.0e62.0 4.0e11.0 <1.0 <1.0 <0.5
K.A. Borges et al. / Energy 67 (2014) 569e574
571
Table 2 Values used at each level of the variables studied during the transesterification of soybean oil in the experimental design. Ethylic transesterification design 1
Ethylic transesterification design 2
Methylic transesterification design 1
Methylic transesterification design 2
Variables
Levels
Levels
Levels
Levels
1
þ1
1
þ1
1
þ1
1
þ1
(1) (2) (3) (4) (5) (6)
30 100 9:1 KOH 0.5% 35
60 400 12:1 NaOH 1.5% 55
30 100 9:1 KOH 0.2% 35
60 400 12:1 KEtox 0.8% 55
30 100 6:1 KOH 0.2% 35
60 400 9:1 NaOH 0.8% 55
30 100 6:1 KOH 0.2% 35
60 400 9:1 KMetox 0.8% 55
Time (min) Stirring (rpm) Molar ratio (alcohol:oil) Catalyst % Catalyst T ( C)
Longer reaction times reduced the yield. As for factor (6) higher temperatures (55 C) led to better yields. Therefore, the equilibrium of the transesterification reaction shifted toward product formation with increasing temperature. Concerning the rotation factor, the transesterification augmented when stirring increased from 100 to 400 rpm, probably because higher rotation promoted better interaction between the reactants. Increasing the alcohol/oil ratio 9:1e12:1 reduced the yield, because an emulsion system was formed. Compared with NaOH, KOH provided better yields: in both cases, the concentration of alkaline catalysts significantly reduced product yield as a result of parallel saponification reactions. Interactions between factors (1) and (6), and (1) and (2) were significant and positive. Therefore, maintaining the individual effects of the factors increased transesterification. Interaction
between factors (2) and (6) was significant and negative, i.e., transesterification decreased. Hence, it is necessary to reverse the effects of these primary factors to improve reaction yield. The Pareto chart also showed that factor (5) (catalyst concentration) was the most significant in the process: the yield decreased when the percentage of catalyst ranged from 0.5% to 1.5% relative to the mass of oil. Factor (4) (catalyst type) was also significant: KOH gave better yields than NaOH. On the basis of statistical analysis, a reaction time of 30 min, rotation at 100 rpm, ethanol/oil molar ratio of 9:1, KOH at 0.5 mass% relative to oil, and temperature of 55 C, furnished the highest yield of ethylic biodiesel from soybean oil. The Pareto chart obtained for the design depicted in Table 4 showed that interaction between the variables (1) and (4), (1) and (5), (2) and (6), and the variables catalyst type (4), time (1), and
Table 3 Experimental design of ethylic transesterification of soybean oil using KOH or NaOH as catalyst. Factor assay
Time (min)
Stirring (rpm)
Molar ratio
Catalyst
Catalyst percentage (%)
Temperature ( C)
Yield (%)
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
30 30 30 60 30 30 30 60 60 30 30 60 60 60 60 60
100 100 100 100 400 400 400 400 100 100 400 100 100 400 400 400
9:1 9:1 12:1 12:1 12:1 9:1 12:1 9:1 9:1 12:1 9:1 9:1 12:1 9:1 12:1 12:1
KOH (1) KOH (1) NaOH (þ1) KOH (1) KOH (1) NaOH (þ1) KOH (1) KOH (1) NaOH (þ1) NaOH (þ1) NaOH (þ1) NaOH (þ1) KOH (1) KOH (1) NaOH (þ1) NaOH (þ1)
1.5 0.5 0.5 1.5 1.5 0.5 0.5 1.5 1.5 1.5 1.5 0.5 0.5 0.5 1.5 0.5
55 35 55 35 55 55 35 35 55 35 35 35 55 55 55 35
57 1 95.7 0.3 98.0 0.4 0.0 43.0 0.4 86 2 97.8 0.2 54.9 0.1 81 1 76.7 0.3 81.6 0.6 96.9 0.9 94.1 0.7 95 1 79.2 0.2 95.6 0.6
(1) (1) (1) (þ1) (1) (1) (1) (þ1) (þ1) (1) (1) (þ1) (þ1) (þ1) (þ1) (þ1)
(1) (1) (1) (1) (þ1) (þ1) (þ1) (þ1) (1) (1) (þ1) (1) (1) (þ1) (þ1) (þ1)
(1) (1) (þ1) (þ1) (þ1) (1) (þ1) (1) (1) (þ1) (1) (1) (þ1) (1) (þ1) (þ1)
(þ1) (1) (1) (þ1) (þ1) (1) (1) (þ1) (þ1) (þ1) (þ1) (1) (1) (1) (þ1) (1)
(þ1) (1) (þ1) (1) (þ1) (þ1) (1) (1) (þ1) (1) (1) (1) (þ1) (þ1) (þ1) (1)
Table 4 Experimental design for the ethylic transesterification of soybean oil employing KOH and potassium ethoxide (EtOK) as catalysts. Factor assay
Time (min)
Stirring (rpm)
Molar ratio
Catalyst
Catalyst percentage (%)
Temperature ( C)
Yield (%)
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
30 60 30 60 30 60 30 60 30 60 30 60 30 60 30 60
100 100 400 400 100 100 400 400 100 100 400 400 100 100 400 400
9:1 9:1 9:1 9:1 12:1 12:1 12:1 12:1 9:1 9:1 9:1 9:1 12:1 12:1 12:1 12:1
EtOK (þ1) EtOK (þ1) EtOK (þ1) EtOK (þ1) EtOK (þ1) EtOK (þ1) EtOK (þ1) EtOK (þ1) KOH (1) KOH (1) KOH (1) KOH (1) KOH (1) KOH (1) KOH (1) KOH (1)
0.20 0.80 0.80 0.20 0.80 0.20 0.20 0.80 0.20 0.80 0.80 0.20 0.80 0.20 0.20 0.80
35 35 55 55 55 55 35 35 55 55 35 35 35 35 55 55
87.4 92.0 91 86.90 90.0 91.8 96.0 91.5 91.2 96.9 79.6 91.0 97.2 95.8 98.6 75
(1) (þ1) (1) (þ1) (1) (þ1) (1) (þ1) (1) (þ1) (1) (þ1) (1) (þ1) (1) (þ1)
(1) (1) (þ1) (þ1) (1) (1) (þ1) (þ1) (1) (1) (þ1) (þ1) (1) (1) (þ1) (þ1)
(1) (1) (1) (1) (þ1) (þ1) (þ1) (þ1) (1) (1) (1) (1) (þ1) (þ1) (þ1) (þ1)
(1) (þ1) (þ1) (1) (þ1) (1) (1) (þ1) (1) (þ1) (þ1) (1) (þ1) (1) (1) (þ1)
(1) (1) (þ1) (þ1) (þ1) (þ1) (1) (1) (þ1) (þ1) (1) (1) (1) (1) (þ1) (þ1)
0.8 0.4 2 0.01 0.8 0.6 0.3 0.2 0.5 0.4 0.6 0.7 0.2 0.2 0.6 2
572
K.A. Borges et al. / Energy 67 (2014) 569e574
Table 5 Experimental design for the methylic transesterification of soybean oil employing KOH or NaOH as catalysts. Factor Time assay (min)
Molar ratio
Catalyst
Temperature Yield (%) Catalyst percentage ( C) (%)
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
6 6 6 6 9 9 9 9 6 6 6 6 9 9 9 9
NaOH (þ1) NaOH (þ1) NaOH (þ1) NaOH (þ1) NaOH (þ1) NaOH (þ1) NaOH (þ1) NaOH (þ1) KOH (1) KOH (1) KOH (1) KOH (1) KOH (1) KOH (1) KOH (1) KOH (1)
0.20 0.80 0.80 0.20 0.80 0.20 0.20 0.80 0.20 0.80 0.80 0.20 0.80 0.20 0.20 0.80
30 60 30 60 30 60 30 60 30 60 30 60 30 60 30 60
(1) (þ1) (1) (þ1) (1) (þ1) (1) (þ1) (1) (þ1) (1) (þ1) (1) (þ1) (1) (þ1)
(1) (1) (1) (1) (þ1) (þ1) (þ1) (þ1) (1) (1) (1) (1) (þ1) (þ1) (þ1) (þ1)
(1) (þ1) (þ1) (1) (þ1) (1) (1) (þ1) (1) (þ1) (þ1) (1) (þ1) (1) (1) (þ1)
35 35 55 55 55 55 35 35 55 55 35 35 35 35 55 55
(1) (1) (þ1) (þ1) (þ1) (þ1) (1) (1) (þ1) (þ1) (1) (1) (1) (1) (þ1) (þ1)
91 75 75 92.0 87.1 91 88 84.8 93.6 90.34 90.00 91 89.9 89 90.8 90.2
2 3 3 0.4 0.2 1 2 0.3 0.3 0.04 0.08 1 0.2 4 0.1 0.2
temperature (6) did not significantly affect transesterification. For this reason, they were disregarded in subsequent statistical analyses. The interactions between variables (1) and (3), (1) and (2), and (1) and (6) were significant and negative, implying that the combined action of two factors, shielded the system from the individual effects of the variables and decreased the reaction yield. Therefore, it was necessary to reverse the trend of interaction, to favor transesterification. For example, carrying out the reaction for long time (60 min) should improve the reaction yield. Interaction between factors (2) and (4) was also significantly positive, so using EtOK should favor the reaction. Increasing the stirring from 100 to 400 rpm increases the yield. Hence, stirring should be conducted at 400 rpm. Regarding the process temperature, 35 C furnished better results, implying that the direct reaction was exothermic and equilibrium shifted toward product formation with decreasing temperature. Larger catalyst concentration (0.8%) reduced the reaction yield, probably due to parallel saponification reactions. On the basis of this statistical analysis, reactions accomplished for 60 min, 35 C, at 400 rpm, and ethanol/oil molar ratio of 12:1, with EtOK 0.2 mass% relative to the oil, afforded the best product yield regarding ethylic biodiesel production from soybean oil. The Pareto chart illustrated for the design presented in Table 5 (Fig. 1c) revealed that the interaction between variables (1) and (4), (2) and (4), (1) and (5), and (1) and (3) was not significant, so they were not considered during statistical analysis. The variables
Table 7 Physical and chemical properties of biodiesel produced in the optimizations I, II, III, IV. Physical properties
Unit
Biodiesel design I
Biodiesel design II
Biodiesel design III
Biodiesel design IV
Density Moisture Acidity Kinematic viscosity Oxidative stability Carbon residue
kg m3 mg kg1 mg KOH g1 mm2 s1 H % mass
870 254.0 0.17 4.3 6.59 0.008
891.8 493.64 0.17 4.4 7.81 0.04
885 327.3 0.02 5.1 7.18 0.002
883 439.3 0.03 4.4 5.33 0.004
rotation (2), molar ratio (3), temperature (6), and time (1) were not significant either. The interactions between (1) and (6) and (1) and (2) were significant and positive; the interaction between (2) and (6) was negative. If the trend of the individual factors is reverted, the effect of the interactions between (1) and (6) and (1) and (2) will be impaired as these interactions are more effective in the yield of the reaction, the interaction (2) and (6) was disregarded. Catalyst concentration (5) and type (4) were the most significant factors in the methylic transesterification of soybean oil. Hence, a KOH concentration of 0.2% relative to the mass furnished the best transesterification results. As for the design presented in Table 6, the Pareto chart (Fig. 1d), showed that interactions between variables (2) and (4) and (2) and (6) were significant and negative. Rotation (2) was less significant than catalyst type (4) and temperature (6). Therefore, it is necessary to reverse the trend of rotation (2) to favor the interaction between (2) and (4) and (2) and (6); a rotation of 400 rpm should be employed. Catalyst concentration (5) and type (4) were the most significant variables in the methylic transesterification of soybean oil. KOH 0.2% gave the best results. The Pareto chart (Fig. 1d) showed that larger values of the variables temperature (6), rotation (2), and alcohol/oil molar ratio (3) should be used. Therefore, methylic transesterification of soybean oil should be carried out for 60 min, at 55 C, and 400 rpm and methanol/oil molar ratio of 9:1, with KOH at a concentration of 0.2 mass% relative to the oil. 3.1.2. Physicochemical characterization of the biodiesels obtained at the optimized conditions After performing the factorial designs, we subjected the esters obtained under the conditions optimized for each design to physicochemical characterization. We compared the values found
Table 6 Factorial design matrix for the soybean oil using methanol as solvent, and KOH or potassium methoxide (MetOK) as catalyst. Factor assay
Time (min)
Stirring (rpm)
Molar ratio (ethanol:oil)
Catalyst
Catalyst percentage (%)
Temperature ( C)
Yield (%)
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
30 60 30 60 30 60 30 60 30 60 30 60 30 60 30 60
100 100 400 400 100 100 400 400 100 100 400 400 100 100 400 400
6 6 6 6 9 9 9 9 6 6 6 6 9 9 9 9
MetOK (þ1) MetOK (þ1) MetOK (þ1) MetOK (þ1) MetOK (þ1) MetOK (þ1) MetOK (þ1) MetOK (þ1) KOH (1) KOH (1) KOH (1) KOH (1) KOH (1) KOH (1) KOH (1) KOH (1)
0.20 0.80 0.80 0.20 0.80 0.20 0.20 0.80 0.20 0.80 0.80 0.20 0.80 0.20 0.20 0.80
35 35 55 55 55 55 35 35 55 55 35 35 35 35 55 55
93.6 92.4 92.6 95.3 95.0 96.6 97 93.6 93.6 90.34 90.00 91 89.9 89 90.8 90.2
(1) (þ1) (1) (þ1) (1) (þ1) (1) (þ1) (1) (þ1) (1) (þ1) (1) (þ1) (1) (þ1)
(1) (1) (þ1) (þ1) (1) (1) (þ1) (þ1) (1) (1) (þ1) (þ1) (1) (1) (þ1) (þ1)
(1) (1) (1) (1) (þ1) (þ1) (þ1) (þ1) (1) (1) (1) (1) (þ1) (þ1) (þ1) (þ1)
(1) (þ1) (þ1) (1) (þ1) (1) (1) (þ1) (1) (þ1) (þ1) (1) (þ1) (1) (1) (þ1)
(1) (1) (þ1) (þ1) (þ1) (þ1) (1) (1) (þ1) (þ1) (1) (1) (1) (1) (þ1) (þ1)
0.4 0.8 0.2 0.5 0.6 0.6 1 0.6 0.3 0.04 0.08 1 0.2 4 0.1 0.2
K.A. Borges et al. / Energy 67 (2014) 569e574
573
Fig. 1. (a) Pareto chart obtained from the experimental design of ethylic soybean oil transesterification comparing the catalysts NaOH and KOH. (b) Pareto chart obtained from the experimental design of ethylic transesterification comparing the catalysts KOH and EtOK. (c) Pareto chart obtained from the experimental design of methylic soybean oil transesterification comparing the catalysts NaOH and KOH. (d) Pareto chart obtained from the experimental design of methylic soybean oil transesterification comparing the catalysts KOH and MetOK.
during these characterizations with the limits established by resolution no 07/2008 of ANP; the biodiesels obtained in the optimized conditions agreed with the recommendations set in the resolution. These biodiesels can be added to the diesel mix sold in the domestic market. Some properties of the biodiesel are related to the molecular structure of its constituent alkyl esters [16], such as density and kinematic viscosity. The density of biodiesel is directly linked to the molecular structure of its components. The larger the length of the carbon chain of the alkyl ester, the higher the density. However, this value decreases upon increasing number of unsaturations in the molecule. The presence of impurities such as alcohol and adulterants also affects the density of the biodiesel. Because density is very sensitive to impurities, the Brazilian ANP resolution states that the biodiesel should be market within one month of certification. After this period, density must be analyzed at 20 C again. If the difference is less than 3.0 kg m3 compared with the certified value, the biodiesel has to be re-analyzed in terms of the water content, the acid content, and the stability to oxidation at 110 C. If the difference is greater than 3.0 kg m3, it is necessary to re-evaluate all quality parameters mentioned in the resolution [29]. The densities of methylic and ethylic biodiesels fulfilled the limits established by ANP (850e900 kg m3) and were close to those reported in the literature [30]: between 869.48 and 880.3 kg m3 for the methylic and ethylic routes, respectively. The viscosity of biodiesel increases with the carbon chain length and the degree of saturation; this parameter influences the combustion process in the combustion chamber of the engine. High viscosity causes heterogeneous biodiesel combustion due to less
efficient atomization in the combustion chamber, culminating in deposition of residues on the internal parts of the engine. Residual soaps, unreacted glycerides (mono-, di-, and triglycerides), and oxidative degradation products of biodiesel increase viscosity. These contaminants can therefore be monitored indirectly by determining the kinematic viscosity at 40 C. ASTM D6751 (test method D 445) provides an acceptable value of viscosity of 3.0e 6.0 mm2 s1. We found that the viscosity of the synthesized methylic and ethylic esters did not fall in this range even though they somewhat differed from literature values [31]: 5.75 for methylic esters and 5.83 for ethylic esters. The oxidative stability of biodiesel is directly related to the degree of unsaturation of the alkyl esters present in the fuel and to the position of the double bond in the carbon chain. The concentration of alkyl esters with high degree of unsaturation varies depending on the raw material used to produce the biodiesel. The greater the number of unsaturations, the more susceptible the molecule is to both thermal and oxidative degradation, forming insoluble products that cause problems of fouling and clogging of the injection system of the engine [29]. The biodiesels analyzed here were oxidatively stable, except for the methylic biodiesel obtained through design IV. The addition of natural antioxidants (e.g., tocopherol) from vegetable oils enhances stability to oxidation; mixing with another biodiesel (blend) of good oxidative stability (in this case, some of the remaining produced biodiesels) also corrects this problem [30]. The monitoring of acidity in biodiesel is very important during storage; alteration in acidity values may reflect the presence of water. The method ASTM D-664 establishes a maximum acidity of
574
K.A. Borges et al. / Energy 67 (2014) 569e574
0.5 mg KOH g1 oil. The biodiesels synthesized here showed high acidity below this limit; the obtained values were similar to those cited by Ferrari et al. [5]. As for the carbon residue, which corresponds to the tendency to form carbonic residues [5], it is an important parameter to analyze, because it correlates with the formation of injector deposits. The methylic and ethylic esters obtained from soybean oil presented values lower than those recommended in the specifications. 4. Conclusions The physico-chemical characterization undertaken to soybean oil was extremely important to indicate the feasibility of transesterification process with this oil, since the quality of biodiesel produced is a direct function of the chemical composition of the oil used as starting material. As for biodiesels produced, only the methyl soy biodiesel produced in the experimental condition described in design IV did not fit the standards of ANP in oxidative stability parameter. The optimization of the ethylic route for the transesterification of soybean oil showed that the addition of 0.2% of EtOK catalyst, molar ratio ethanol:Oil 12:1, with a reaction time of 60 min, temperature of 35 C, and stirring at 100 rpm, provides yields above 90%. However when using the KOH catalyst, which has a lower cost compared to potassium ethoxide, it was possible to obtain satisfactory yields (above 85%) and even in this condition, are required minor amounts of reagents and shorter reaction time. The methylic route optimization showed that the ideal catalyst for transesterification is KOH. The optimal temperature for obtaining higher yields is to 55 C, and the schedules shown that the use of reactants at a ratio of 6:1, at 100 rpm, catalyst concentration 0.2 mass% relative to oil, time 30 min and provides 93% yield. The work performed allowed the best condition for the production of biodiesel from soybean oil by methylic and ethylic routes. To obtain the best reaction conditions as well as taking into account income, was also considered that the costs imposed on the income process. It is noteworthy that soybean oil meets the need for real Brazilian biodiesel production, once it is an abundant raw material, inexpensive and suitable chemical composition and, moreover, its biodiesel meets the quality parameters in terms of physico-chemical properties. Acknowledgments The authors are grateful to Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP) and Fundação de Amparo à Pesquisa do Estado de Minas Gerais (FAPEMIG) for financial support. Additionally, the authors would like to thank Dr. Cynthia Maria de Campos Prado Manso for revising and editing the text. References [1] Suarez PAZ, Meneghetti SMP. 70th Anniversary of biodiesel in 2007: historical evolution and current situation in Brazil. Quim Nova 2007;30:2068e71. [2] Pousa GPAG, Santos ALF, Suarez PAZ. History and policy of biodiesel in Brazil. Energy Policy 2007;35:5393e8. [3] Rinaldi R, Garcia C, Marciniuk LL, Rossi AV, Schuchardt U. Synthesis of biodiesel: a contextualized experiment proposal for the general chemistry laboratory. Quim Nova 2007;30:1374e80.
[4] Demirbas A. Energy and environmental issues relating to greenhouse gas emissions in Turkey. Energy Convers Manage 2003;44:203e13. [5] Ferrari RA, Oliveira VD, Scabio A. Biodiesel from soybean: characterization and consumption in an energy generator. Quim Nova 2005;28:19e23. [6] Lotero E, Liu YJ, Lopez DE, Suwannakarn K, Bruce DA, Goodwin JG. Synthesis of biodiesel via acid catalysis. Ind Eng Chem Res 2005;44:5353e63. [7] Oliveira JS, Montalvao R, Daher L, Suarez PAZ, Rubim JC. Determination of methyl ester contents in biodiesel blends by FTIR-ATR and FTNIR spectroscopies. Talanta 2006;69:1278e84. [8] Correa SM, Arbilla G. Aromatic hydrocarbons emissions in diesel and biodiesel exhaust. Atmos Environ 2006;40:6821e6. [9] Nguyen NT, Demirel Y. A novel biodiesel and glycerol carbonate production plant. Int J Chem React Eng 2011;9:1e25. [10] Vyas AP, Verma JL, Subrahmanyam N. A review on FAME production processes. Fuel 2010;89:1e9. [11] Abbaszaadeh A, Ghobadian B, Omidkhah MR, Naja&filig G. Current biodiesel production technologies: a comparative review. Energy Convers Manage 2012;63:138e48. [12] Yun H, Wang M, Feng W, Tan T. Process simulation and energy optimization of the enzyme-catalyzed biodiesel production. Energy 2013;54:84e96. [13] Singh SP, Singh Dipti. Biodiesel production through the use of different sources and characterization of oils and their esters as the substitute of diesel: a review. Renew Sustain Energy Rev 2010;14:200e16. [14] Serrano M, Oliveros R, Sánchez M, Moraschini A, Martínez M, Aracil J. Influence of blending vegetable oil methyl esters on biodiesel fuel properties: oxidative stability and cold flow properties. Energy 2013;65:109e15. [15] Crabbe E, Nolasco-Hipolito C, Kobayashi G, Sonomoto K, Ishizaki A. Biodiesel production from crude palm oil and evaluation of butanol extraction and fuel properties. Process Biochem 2001;37:65e71. [16] Demirbas A. Comparison of transesterification methods for production of biodiesel from vegetable oils and fats. Energy Convers Manage 2008;49: 125e30. [17] Berrios M, Gutiérrez MC, Martín MA, Martín A. Application of the factorial design of experiments to biodiesel production from lard. Fuel Process Technol 2009;90:1447e51. [18] Hájek M, Skopal F, Capek L, Cernoch M, Kutálek P. Ethanolysis of rapeseed oil by KOH as homogeneous and as heterogeneous catalyst supported on alumina and CaO. Energy 2012;48:392e7. [19] Kiliç M, Uzun BB, Pütün E, Pütün AE. Optimization of biodiesel production from castor oil using factorial design. Fuel Process Technol 2013;111: 105e8. [20] Veli ckovi c AV, Stamenkovi c OS, Todorovi c ZB, Veljkovi&cacute VB. Application of the full factorial design to optimization of base-catalyzed sunflower oil ethanolysis. Fuel 2013;104:433e42. [21] Mendonça DR, Andrade HMC, Guimarães PRB, Vianna RF, Meneghetti SMP, Pontes LAM, et al. Application of full factorial design and Doehlert matrix for the optimisation of beef tallow methanolysis via homogeneous catalysis. Fuel Process Technol 2011;92:342e8. [22] Rashid U, Rehman HA, Hussain I, Ibrahim M, Haider MS. Muskmelon (Cucumis melo) seed oil: a potential non-food oil source for biodiesel production. Energy 2011;36:5632e9. [23] Allawzi M, Kandah MI. Parametric study of biodiesel production from used soybean oil. Eur J Lipid Sci Technol 2008;110:760e7. [24] Noureddini H, Zhu D. Kinetics of transesterification of soybean oil. JAOCS 1997;74:1457e63. [25] Oliveira D, Luccio MD, Faccio C, Rosa CD, Bender JP, Lipke N, et al. Transesterification of soybean oil and castor oil for biodiesel production. Appl Biochem Biotechnol 2005;121:553e60. [26] Firestone DE. Official methods and recommended practices of The American Oil Chemists Society. Illinois: AOCS Press; 1997. [27] Bondioli P, Della Bella L. An alternative spectrophotometric method for the determination of free glycerol in biodiesel. Eur J Lipid Sci Technol 2005;107: 153e7. [28] Sharma YC, Singh B, Upadhyay SN. Advancements in development and characterization of biodiesel: a review. Fuel 2008;87:2355e73. [29] Lobo IP, Ferreira SLC, da Cruz RS. Biodiesel: quality parameters and analytical methods. Quim Nova 2009;32:1596e608. [30] Santos AGD. Avaliação da estabilidade térmica e oxidativa do biodiesel de algodão, girassol, dendê e sebo bovino. Natal-RN, 183 p. (Tese) e Universidade Federal do Rio Grande do Norte. [31] Candeia RA. Biodiesel de Soja: Síntese, Degradação e Misturas Binárias. João Pessoa- PB, 150 p. (Tese). Universidade Federal da Paraíba; 2008.
Contents lists available at ScienceDirect
Energy journal homepage: www.elsevier.com/locate/energy
Homogeneous catalysis of soybean oil transesterification via methylic and ethylic routes: Multivariate comparison Karen Araújo Borges a, André Luiz Squissato a, Douglas Queiroz Santos a, Waldomiro Borges Neto a, Antônio Carlos Ferreira Batista b, Tiago Almeida Silva c, Andressa Tironi Vieira d, Marcelo Firmino de Oliveira d, *, Manuel Gonzalo Hernández-Terrones a a
Instituto de Química, Universidade Federal de Uberlândia, Campus Santa Mônica, 38400-902, Uberlândia, Minas Gerais, Brazil Laboratório de Energias Renováveis e Meio ambiente do Pontal (LERMAP), Universidade Federal de Uberlândia, 38302-000, Ituiutaba, MG, Brazil c Departamento de Química, Universidade Federal de São Carlos, 13560-970, São Carlos, SP, Brazil d Departamento de Química, Faculdade de Filosofia, Ciências e Letras de Ribeirão Preto, Universidade de São Paulo, 14040-901, Ribeirão Preto, SP, Brazil b
a r t i c l e i n f o
a b s t r a c t
Article history: Received 21 September 2013 Received in revised form 31 January 2014 Accepted 2 February 2014 Available online 28 February 2014
An experiment to establish the best reaction conditions for the transesterification of soybean oil is described. We conducted the ethylic and methylic routes using two different protocols, and evaluated how the variables time, stirring, alcohol/oil molar ratio, catalyst (%), catalyst type, and temperature affected the process. The highest yield of biodiesel was obtained using the following conditions: ethylic route e t ¼ 60 min, stirring: 100 rpm, ethanol/oil molar ratio ¼ 12:1, catalyst relative to oil (%) ¼ 0.2%, catalyst ¼ potassium ethoxide, temperature ¼ 35 C; methylic route e t ¼ 30 min, stirring: 100 rpm, methanol/oil molar ratio ¼ 6:1, catalyst (%) ¼ 0.2%, catalyst ¼ KOH, temperature ¼ 55 C. We analyzed the acidity, moisture content, density at 20 C, kinematic viscosity at 40 C, oxidative stability, and carbon residue at the biodiesels obtained under optimal conditions. The results were consistent with the values allowed by the Brazilian ANP (Resolution 07/2008). We also conducted the physicochemical analysis of the soybean oil used as feedstock to produce biodiesel. Ó 2014 Elsevier Ltd. All rights reserved.
Keywords: Transesterification Soybean oil Optimization Biodiesel
1. Introduction The oil crisis seen in recent decades together with growing demand for fuel and increasing concern about the environment has prompted the search for alternative energy sources both in Brazil and worldwide [1,2]. Research has focused on developing new basic inputs of renewable character, to produce fuels that can replace petroleum products. In this context, biomass plays a leading role. It is renewable, widely available, biodegradable, and inexpensive [1,2]. Indeed, plant oils had already been tested and used as fuel in diesel engines before the advent of petroleum diesel. However, for both economic and technical reasons, they gave way to diesel oil [3,4]. Currently, biodiesel is a well established example of the use of biomass to produce energy and offers advantages over petroleum diesel [5e7]. It is non-toxic, originates from renewable sources, and
* Corresponding author. Tel./fax: þ55 34 3239 4425. E-mail addresses: [email protected], [email protected], fl[email protected] (M.F. de Oliveira). http://dx.doi.org/10.1016/j.energy.2014.02.012 0360-5442/Ó 2014 Elsevier Ltd. All rights reserved.
leads to better quality of emissions during the combustion process [8]. Although biodiesel provides about 10% less energy than diesel fuel, motor performance is essentially the same with respect to power and torque [6]. Additionally, biodiesel is highly viscous, enhancing lubricity and reducing wear of the moving parts of the engine. Several processes exist to produce biodiesel, but transesterification is the method that is often employed most worldwide [9e11]. It involves reacting a lipid (known as triglycerides or triacylglycerols) with a mono-short chain alcohol (methyl or ethyl) in the presence of a catalyst (acid or base), which produces a mixture of alkyl esters of fatty acids (known as biodiesel) and glycerol [9e11]. The transesterification reaction has been conduced employing different catalytic routes. The catalysts used for the transesterification can be classified into: homogeneous and heterogeneous (basic and acid), or biological (enzymes) [9,12,13]. Among all catalytic routes, the basic homogeneous catalysts, typically sodium and potassium hydroxide, in fact have been most often used industrially, by various as reasons, such as [9,12e14]: low cost catalyst, high catalytic activity with maximum conversion achieved
570
K.A. Borges et al. / Energy 67 (2014) 569e574
in a minimal time, simple operational due to the mild conditions, and minor problem with the corrosive effects. Many factors affect the transesterification of oils and fats; for example, the presence of free fatty acids, humidity, alcohol type, molar alcohol/oil ratio, catalyst type and concentration, reaction time, and temperature [15,16]. Therefore, studies involving transesterification of vegetable oil must be performed considering the effect of the various factors during the optimization, and simultaneously, because the experimental parameters are interrelated. This necessity/approach can be achieved from the use factorial design, where from a small number of experiments diverse variables are studies simultaneously at different levels and taking into account its interactions with the others variables [17,18]. Thus, the recent literature reports the application of factorial design for the optimization of the homogeneous catalysis transesterification from different feedstocks, such as lard [17], castor oil [19], sunflower oil [20], beef tallow [21], Muskmelon (Cucumis melo) seed oil [22], waste vegetable oil [23], among others. In regard the biodiesel production from soybean oil the traditional univariate methodology for optimization of the process has been adopted [24]. However, Oliveira et al. [25] presented the optimization of alkaline transesterification of soybean oil using a Taguchi experimental design, and from this experimental design the effect of the variables temperature, reaction time, catalyst concentration and oil-toethanol molar ratio were available. The influence of other very important factor were not considered in this reported work, as type of alcohol and basic catalyst [25]. In this work a novel approach for the homogeneous catalysis optimization of soybean oil transesterification via methylic and ethylic routes is presented, which all experimental parameters were systematically evaluated using a factorial design. Thus we investigated which of the variables reaction time, reaction rate, molar alcohol/oil ratio, catalyst type, catalyst concentration, and temperature significantly influence the yield of the transesterification reaction using a fractional factorial design 262 in duplicate, which resulted in 32 experiments conducted for each design. 2. Experimental 2.1. Characterization of the soybean oil and biodiesel Soybean oil was analyzed for the following parameters: refractive index (40 C), saponification index, acidity, moisture content, density at 20.0 C, kinematic viscosity at 40 C; oxidative stability, and carbon residue. The same analyses were also conducted for the biodiesel obtained in the optimum experimental conditions, except for the refractive index and saponification number. The parameters refractive index (40 C), saponification number, and acidity were determined according to the official procedures recommended by the American Oil Chemists Society [26]. The moisture content was analyzed according to the norm ASTM D-6304 using a Karl Fischer colorimetric titrator model 831 KF. The density was determined according to the norm ASTM D-4052, which corresponds to the Brazilian norm ABNT NBR 14065, using a DA-500-Kyoto densimeter. Kinematic viscosity at 40 C was obtained according to the norms ASTM D-445 and ASTM 446. The oxidative stability was analyzed by the method EN 14112, on a Rancimat equipment model 743 from Metrohm. 2.2. Transesterification of the soybean oil Refined commercial soybean oil, household, packaging 900 mL, model Liza, industrially processed by Cargill Agricultural SA was used. Table 1 shows the fatty acid composition of the oil.
For the transesterification reaction, the catalyst was added to the solvent; the mixture was stirred until complete dissolution and transferred to a 250 mL Erlenmeyer flask. Next, the oil was added, and transesterification was performed by following the experimental conditions described in the experimental design ethyl 1 and 2 and methyl 1 and 2 (Table 2). After transesterification, two phases were noted: upper and lower crude biodiesel glycerin. Then, biodiesel was removed from the mixture and washed five times with water at 80 C at a water/biodiesel ratio 1:3 (v/v), to remove impurities. The biodiesel was then dried in a rotary evaporator. The wash water containing the residue catalyst was properly neutralized with 0.1 mol L1 HCl solution before being discarded. 2.3. Experimental design To establish the experimental conditions that improved the efficiency of transesterification, the reaction was carried out using a fractional factorial experimental design, in duplicate. The option used in the experiment was 262, which is interesting for investigative purposes, because it reduces the number of tests. This design is not completely saturated and no main effects mix with first order, which ensures that the effects of the variables analyzed in the response are reliably calculated, without statistical information quality. Table 2 lists the values employed at each level of the variables, chosen on the basis of literature studies [27,28]. 3. Results and discussion 3.1. Experimental design 3.1.1. Optimization of the ethylic route using KOH or NaOH as catalyst Tables 3e6 present the corresponding values of mass production regarding the methylic and ethylic transesterification of soybean oil conducted according to the experimental design 262. We compared the performance of the catalysts KOH, NaOH, and potassium ethoxide in the ethylic route of soybean oil transesterification. We also compared the use of KOH, NaOH, and potassium methoxide as catalysts in the methylic route. We calculated the mass yield of the process on the basis of the transesterification reaction stoichiometry: 1 mol of oil affords 3 mol of ester (Table 7). Fig. 1 represents the Pareto chart obtained from the data summarized in Tables 3e6 using the fractional factorial design; we evaluated the effects of each variable and their interactions on the yield of soybean oil transesterification. The Pareto chart corresponding to the design shown in Table 3 (Fig. 1a) revealed that the interactions between factors (1) and (4), (1) and (3), (2) and (5), (1) and (5) affected transesterification less than the variables disregarded then in subsequent statistical analyses.
Table 1 Composition of fatty acids in soybean. Fatty acids
Structure
Reference values (%)
e Myristic Palmitic Palmitoleic Estearic Oleic (Omega 9) Linoleic (Omega 6) Linolenic (Omega 3) Araquidic Eicosenoic Behenic
C<14 C14:0 C16:0 C16:1 C18:0 C18:1 C18:2 C18:3 C20:0 C20:1 C22:0
<0.1 <0.5 7.0e14.0 <0.5 1.4e5.5 19.0e30.0 44.0e62.0 4.0e11.0 <1.0 <1.0 <0.5
K.A. Borges et al. / Energy 67 (2014) 569e574
571
Table 2 Values used at each level of the variables studied during the transesterification of soybean oil in the experimental design. Ethylic transesterification design 1
Ethylic transesterification design 2
Methylic transesterification design 1
Methylic transesterification design 2
Variables
Levels
Levels
Levels
Levels
1
þ1
1
þ1
1
þ1
1
þ1
(1) (2) (3) (4) (5) (6)
30 100 9:1 KOH 0.5% 35
60 400 12:1 NaOH 1.5% 55
30 100 9:1 KOH 0.2% 35
60 400 12:1 KEtox 0.8% 55
30 100 6:1 KOH 0.2% 35
60 400 9:1 NaOH 0.8% 55
30 100 6:1 KOH 0.2% 35
60 400 9:1 KMetox 0.8% 55
Time (min) Stirring (rpm) Molar ratio (alcohol:oil) Catalyst % Catalyst T ( C)
Longer reaction times reduced the yield. As for factor (6) higher temperatures (55 C) led to better yields. Therefore, the equilibrium of the transesterification reaction shifted toward product formation with increasing temperature. Concerning the rotation factor, the transesterification augmented when stirring increased from 100 to 400 rpm, probably because higher rotation promoted better interaction between the reactants. Increasing the alcohol/oil ratio 9:1e12:1 reduced the yield, because an emulsion system was formed. Compared with NaOH, KOH provided better yields: in both cases, the concentration of alkaline catalysts significantly reduced product yield as a result of parallel saponification reactions. Interactions between factors (1) and (6), and (1) and (2) were significant and positive. Therefore, maintaining the individual effects of the factors increased transesterification. Interaction
between factors (2) and (6) was significant and negative, i.e., transesterification decreased. Hence, it is necessary to reverse the effects of these primary factors to improve reaction yield. The Pareto chart also showed that factor (5) (catalyst concentration) was the most significant in the process: the yield decreased when the percentage of catalyst ranged from 0.5% to 1.5% relative to the mass of oil. Factor (4) (catalyst type) was also significant: KOH gave better yields than NaOH. On the basis of statistical analysis, a reaction time of 30 min, rotation at 100 rpm, ethanol/oil molar ratio of 9:1, KOH at 0.5 mass% relative to oil, and temperature of 55 C, furnished the highest yield of ethylic biodiesel from soybean oil. The Pareto chart obtained for the design depicted in Table 4 showed that interaction between the variables (1) and (4), (1) and (5), (2) and (6), and the variables catalyst type (4), time (1), and
Table 3 Experimental design of ethylic transesterification of soybean oil using KOH or NaOH as catalyst. Factor assay
Time (min)
Stirring (rpm)
Molar ratio
Catalyst
Catalyst percentage (%)
Temperature ( C)
Yield (%)
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
30 30 30 60 30 30 30 60 60 30 30 60 60 60 60 60
100 100 100 100 400 400 400 400 100 100 400 100 100 400 400 400
9:1 9:1 12:1 12:1 12:1 9:1 12:1 9:1 9:1 12:1 9:1 9:1 12:1 9:1 12:1 12:1
KOH (1) KOH (1) NaOH (þ1) KOH (1) KOH (1) NaOH (þ1) KOH (1) KOH (1) NaOH (þ1) NaOH (þ1) NaOH (þ1) NaOH (þ1) KOH (1) KOH (1) NaOH (þ1) NaOH (þ1)
1.5 0.5 0.5 1.5 1.5 0.5 0.5 1.5 1.5 1.5 1.5 0.5 0.5 0.5 1.5 0.5
55 35 55 35 55 55 35 35 55 35 35 35 55 55 55 35
57 1 95.7 0.3 98.0 0.4 0.0 43.0 0.4 86 2 97.8 0.2 54.9 0.1 81 1 76.7 0.3 81.6 0.6 96.9 0.9 94.1 0.7 95 1 79.2 0.2 95.6 0.6
(1) (1) (1) (þ1) (1) (1) (1) (þ1) (þ1) (1) (1) (þ1) (þ1) (þ1) (þ1) (þ1)
(1) (1) (1) (1) (þ1) (þ1) (þ1) (þ1) (1) (1) (þ1) (1) (1) (þ1) (þ1) (þ1)
(1) (1) (þ1) (þ1) (þ1) (1) (þ1) (1) (1) (þ1) (1) (1) (þ1) (1) (þ1) (þ1)
(þ1) (1) (1) (þ1) (þ1) (1) (1) (þ1) (þ1) (þ1) (þ1) (1) (1) (1) (þ1) (1)
(þ1) (1) (þ1) (1) (þ1) (þ1) (1) (1) (þ1) (1) (1) (1) (þ1) (þ1) (þ1) (1)
Table 4 Experimental design for the ethylic transesterification of soybean oil employing KOH and potassium ethoxide (EtOK) as catalysts. Factor assay
Time (min)
Stirring (rpm)
Molar ratio
Catalyst
Catalyst percentage (%)
Temperature ( C)
Yield (%)
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
30 60 30 60 30 60 30 60 30 60 30 60 30 60 30 60
100 100 400 400 100 100 400 400 100 100 400 400 100 100 400 400
9:1 9:1 9:1 9:1 12:1 12:1 12:1 12:1 9:1 9:1 9:1 9:1 12:1 12:1 12:1 12:1
EtOK (þ1) EtOK (þ1) EtOK (þ1) EtOK (þ1) EtOK (þ1) EtOK (þ1) EtOK (þ1) EtOK (þ1) KOH (1) KOH (1) KOH (1) KOH (1) KOH (1) KOH (1) KOH (1) KOH (1)
0.20 0.80 0.80 0.20 0.80 0.20 0.20 0.80 0.20 0.80 0.80 0.20 0.80 0.20 0.20 0.80
35 35 55 55 55 55 35 35 55 55 35 35 35 35 55 55
87.4 92.0 91 86.90 90.0 91.8 96.0 91.5 91.2 96.9 79.6 91.0 97.2 95.8 98.6 75
(1) (þ1) (1) (þ1) (1) (þ1) (1) (þ1) (1) (þ1) (1) (þ1) (1) (þ1) (1) (þ1)
(1) (1) (þ1) (þ1) (1) (1) (þ1) (þ1) (1) (1) (þ1) (þ1) (1) (1) (þ1) (þ1)
(1) (1) (1) (1) (þ1) (þ1) (þ1) (þ1) (1) (1) (1) (1) (þ1) (þ1) (þ1) (þ1)
(1) (þ1) (þ1) (1) (þ1) (1) (1) (þ1) (1) (þ1) (þ1) (1) (þ1) (1) (1) (þ1)
(1) (1) (þ1) (þ1) (þ1) (þ1) (1) (1) (þ1) (þ1) (1) (1) (1) (1) (þ1) (þ1)
0.8 0.4 2 0.01 0.8 0.6 0.3 0.2 0.5 0.4 0.6 0.7 0.2 0.2 0.6 2
572
K.A. Borges et al. / Energy 67 (2014) 569e574
Table 5 Experimental design for the methylic transesterification of soybean oil employing KOH or NaOH as catalysts. Factor Time assay (min)
Molar ratio
Catalyst
Temperature Yield (%) Catalyst percentage ( C) (%)
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
6 6 6 6 9 9 9 9 6 6 6 6 9 9 9 9
NaOH (þ1) NaOH (þ1) NaOH (þ1) NaOH (þ1) NaOH (þ1) NaOH (þ1) NaOH (þ1) NaOH (þ1) KOH (1) KOH (1) KOH (1) KOH (1) KOH (1) KOH (1) KOH (1) KOH (1)
0.20 0.80 0.80 0.20 0.80 0.20 0.20 0.80 0.20 0.80 0.80 0.20 0.80 0.20 0.20 0.80
30 60 30 60 30 60 30 60 30 60 30 60 30 60 30 60
(1) (þ1) (1) (þ1) (1) (þ1) (1) (þ1) (1) (þ1) (1) (þ1) (1) (þ1) (1) (þ1)
(1) (1) (1) (1) (þ1) (þ1) (þ1) (þ1) (1) (1) (1) (1) (þ1) (þ1) (þ1) (þ1)
(1) (þ1) (þ1) (1) (þ1) (1) (1) (þ1) (1) (þ1) (þ1) (1) (þ1) (1) (1) (þ1)
35 35 55 55 55 55 35 35 55 55 35 35 35 35 55 55
(1) (1) (þ1) (þ1) (þ1) (þ1) (1) (1) (þ1) (þ1) (1) (1) (1) (1) (þ1) (þ1)
91 75 75 92.0 87.1 91 88 84.8 93.6 90.34 90.00 91 89.9 89 90.8 90.2
2 3 3 0.4 0.2 1 2 0.3 0.3 0.04 0.08 1 0.2 4 0.1 0.2
temperature (6) did not significantly affect transesterification. For this reason, they were disregarded in subsequent statistical analyses. The interactions between variables (1) and (3), (1) and (2), and (1) and (6) were significant and negative, implying that the combined action of two factors, shielded the system from the individual effects of the variables and decreased the reaction yield. Therefore, it was necessary to reverse the trend of interaction, to favor transesterification. For example, carrying out the reaction for long time (60 min) should improve the reaction yield. Interaction between factors (2) and (4) was also significantly positive, so using EtOK should favor the reaction. Increasing the stirring from 100 to 400 rpm increases the yield. Hence, stirring should be conducted at 400 rpm. Regarding the process temperature, 35 C furnished better results, implying that the direct reaction was exothermic and equilibrium shifted toward product formation with decreasing temperature. Larger catalyst concentration (0.8%) reduced the reaction yield, probably due to parallel saponification reactions. On the basis of this statistical analysis, reactions accomplished for 60 min, 35 C, at 400 rpm, and ethanol/oil molar ratio of 12:1, with EtOK 0.2 mass% relative to the oil, afforded the best product yield regarding ethylic biodiesel production from soybean oil. The Pareto chart illustrated for the design presented in Table 5 (Fig. 1c) revealed that the interaction between variables (1) and (4), (2) and (4), (1) and (5), and (1) and (3) was not significant, so they were not considered during statistical analysis. The variables
Table 7 Physical and chemical properties of biodiesel produced in the optimizations I, II, III, IV. Physical properties
Unit
Biodiesel design I
Biodiesel design II
Biodiesel design III
Biodiesel design IV
Density Moisture Acidity Kinematic viscosity Oxidative stability Carbon residue
kg m3 mg kg1 mg KOH g1 mm2 s1 H % mass
870 254.0 0.17 4.3 6.59 0.008
891.8 493.64 0.17 4.4 7.81 0.04
885 327.3 0.02 5.1 7.18 0.002
883 439.3 0.03 4.4 5.33 0.004
rotation (2), molar ratio (3), temperature (6), and time (1) were not significant either. The interactions between (1) and (6) and (1) and (2) were significant and positive; the interaction between (2) and (6) was negative. If the trend of the individual factors is reverted, the effect of the interactions between (1) and (6) and (1) and (2) will be impaired as these interactions are more effective in the yield of the reaction, the interaction (2) and (6) was disregarded. Catalyst concentration (5) and type (4) were the most significant factors in the methylic transesterification of soybean oil. Hence, a KOH concentration of 0.2% relative to the mass furnished the best transesterification results. As for the design presented in Table 6, the Pareto chart (Fig. 1d), showed that interactions between variables (2) and (4) and (2) and (6) were significant and negative. Rotation (2) was less significant than catalyst type (4) and temperature (6). Therefore, it is necessary to reverse the trend of rotation (2) to favor the interaction between (2) and (4) and (2) and (6); a rotation of 400 rpm should be employed. Catalyst concentration (5) and type (4) were the most significant variables in the methylic transesterification of soybean oil. KOH 0.2% gave the best results. The Pareto chart (Fig. 1d) showed that larger values of the variables temperature (6), rotation (2), and alcohol/oil molar ratio (3) should be used. Therefore, methylic transesterification of soybean oil should be carried out for 60 min, at 55 C, and 400 rpm and methanol/oil molar ratio of 9:1, with KOH at a concentration of 0.2 mass% relative to the oil. 3.1.2. Physicochemical characterization of the biodiesels obtained at the optimized conditions After performing the factorial designs, we subjected the esters obtained under the conditions optimized for each design to physicochemical characterization. We compared the values found
Table 6 Factorial design matrix for the soybean oil using methanol as solvent, and KOH or potassium methoxide (MetOK) as catalyst. Factor assay
Time (min)
Stirring (rpm)
Molar ratio (ethanol:oil)
Catalyst
Catalyst percentage (%)
Temperature ( C)
Yield (%)
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
30 60 30 60 30 60 30 60 30 60 30 60 30 60 30 60
100 100 400 400 100 100 400 400 100 100 400 400 100 100 400 400
6 6 6 6 9 9 9 9 6 6 6 6 9 9 9 9
MetOK (þ1) MetOK (þ1) MetOK (þ1) MetOK (þ1) MetOK (þ1) MetOK (þ1) MetOK (þ1) MetOK (þ1) KOH (1) KOH (1) KOH (1) KOH (1) KOH (1) KOH (1) KOH (1) KOH (1)
0.20 0.80 0.80 0.20 0.80 0.20 0.20 0.80 0.20 0.80 0.80 0.20 0.80 0.20 0.20 0.80
35 35 55 55 55 55 35 35 55 55 35 35 35 35 55 55
93.6 92.4 92.6 95.3 95.0 96.6 97 93.6 93.6 90.34 90.00 91 89.9 89 90.8 90.2
(1) (þ1) (1) (þ1) (1) (þ1) (1) (þ1) (1) (þ1) (1) (þ1) (1) (þ1) (1) (þ1)
(1) (1) (þ1) (þ1) (1) (1) (þ1) (þ1) (1) (1) (þ1) (þ1) (1) (1) (þ1) (þ1)
(1) (1) (1) (1) (þ1) (þ1) (þ1) (þ1) (1) (1) (1) (1) (þ1) (þ1) (þ1) (þ1)
(1) (þ1) (þ1) (1) (þ1) (1) (1) (þ1) (1) (þ1) (þ1) (1) (þ1) (1) (1) (þ1)
(1) (1) (þ1) (þ1) (þ1) (þ1) (1) (1) (þ1) (þ1) (1) (1) (1) (1) (þ1) (þ1)
0.4 0.8 0.2 0.5 0.6 0.6 1 0.6 0.3 0.04 0.08 1 0.2 4 0.1 0.2
K.A. Borges et al. / Energy 67 (2014) 569e574
573
Fig. 1. (a) Pareto chart obtained from the experimental design of ethylic soybean oil transesterification comparing the catalysts NaOH and KOH. (b) Pareto chart obtained from the experimental design of ethylic transesterification comparing the catalysts KOH and EtOK. (c) Pareto chart obtained from the experimental design of methylic soybean oil transesterification comparing the catalysts NaOH and KOH. (d) Pareto chart obtained from the experimental design of methylic soybean oil transesterification comparing the catalysts KOH and MetOK.
during these characterizations with the limits established by resolution no 07/2008 of ANP; the biodiesels obtained in the optimized conditions agreed with the recommendations set in the resolution. These biodiesels can be added to the diesel mix sold in the domestic market. Some properties of the biodiesel are related to the molecular structure of its constituent alkyl esters [16], such as density and kinematic viscosity. The density of biodiesel is directly linked to the molecular structure of its components. The larger the length of the carbon chain of the alkyl ester, the higher the density. However, this value decreases upon increasing number of unsaturations in the molecule. The presence of impurities such as alcohol and adulterants also affects the density of the biodiesel. Because density is very sensitive to impurities, the Brazilian ANP resolution states that the biodiesel should be market within one month of certification. After this period, density must be analyzed at 20 C again. If the difference is less than 3.0 kg m3 compared with the certified value, the biodiesel has to be re-analyzed in terms of the water content, the acid content, and the stability to oxidation at 110 C. If the difference is greater than 3.0 kg m3, it is necessary to re-evaluate all quality parameters mentioned in the resolution [29]. The densities of methylic and ethylic biodiesels fulfilled the limits established by ANP (850e900 kg m3) and were close to those reported in the literature [30]: between 869.48 and 880.3 kg m3 for the methylic and ethylic routes, respectively. The viscosity of biodiesel increases with the carbon chain length and the degree of saturation; this parameter influences the combustion process in the combustion chamber of the engine. High viscosity causes heterogeneous biodiesel combustion due to less
efficient atomization in the combustion chamber, culminating in deposition of residues on the internal parts of the engine. Residual soaps, unreacted glycerides (mono-, di-, and triglycerides), and oxidative degradation products of biodiesel increase viscosity. These contaminants can therefore be monitored indirectly by determining the kinematic viscosity at 40 C. ASTM D6751 (test method D 445) provides an acceptable value of viscosity of 3.0e 6.0 mm2 s1. We found that the viscosity of the synthesized methylic and ethylic esters did not fall in this range even though they somewhat differed from literature values [31]: 5.75 for methylic esters and 5.83 for ethylic esters. The oxidative stability of biodiesel is directly related to the degree of unsaturation of the alkyl esters present in the fuel and to the position of the double bond in the carbon chain. The concentration of alkyl esters with high degree of unsaturation varies depending on the raw material used to produce the biodiesel. The greater the number of unsaturations, the more susceptible the molecule is to both thermal and oxidative degradation, forming insoluble products that cause problems of fouling and clogging of the injection system of the engine [29]. The biodiesels analyzed here were oxidatively stable, except for the methylic biodiesel obtained through design IV. The addition of natural antioxidants (e.g., tocopherol) from vegetable oils enhances stability to oxidation; mixing with another biodiesel (blend) of good oxidative stability (in this case, some of the remaining produced biodiesels) also corrects this problem [30]. The monitoring of acidity in biodiesel is very important during storage; alteration in acidity values may reflect the presence of water. The method ASTM D-664 establishes a maximum acidity of
574
K.A. Borges et al. / Energy 67 (2014) 569e574
0.5 mg KOH g1 oil. The biodiesels synthesized here showed high acidity below this limit; the obtained values were similar to those cited by Ferrari et al. [5]. As for the carbon residue, which corresponds to the tendency to form carbonic residues [5], it is an important parameter to analyze, because it correlates with the formation of injector deposits. The methylic and ethylic esters obtained from soybean oil presented values lower than those recommended in the specifications. 4. Conclusions The physico-chemical characterization undertaken to soybean oil was extremely important to indicate the feasibility of transesterification process with this oil, since the quality of biodiesel produced is a direct function of the chemical composition of the oil used as starting material. As for biodiesels produced, only the methyl soy biodiesel produced in the experimental condition described in design IV did not fit the standards of ANP in oxidative stability parameter. The optimization of the ethylic route for the transesterification of soybean oil showed that the addition of 0.2% of EtOK catalyst, molar ratio ethanol:Oil 12:1, with a reaction time of 60 min, temperature of 35 C, and stirring at 100 rpm, provides yields above 90%. However when using the KOH catalyst, which has a lower cost compared to potassium ethoxide, it was possible to obtain satisfactory yields (above 85%) and even in this condition, are required minor amounts of reagents and shorter reaction time. The methylic route optimization showed that the ideal catalyst for transesterification is KOH. The optimal temperature for obtaining higher yields is to 55 C, and the schedules shown that the use of reactants at a ratio of 6:1, at 100 rpm, catalyst concentration 0.2 mass% relative to oil, time 30 min and provides 93% yield. The work performed allowed the best condition for the production of biodiesel from soybean oil by methylic and ethylic routes. To obtain the best reaction conditions as well as taking into account income, was also considered that the costs imposed on the income process. It is noteworthy that soybean oil meets the need for real Brazilian biodiesel production, once it is an abundant raw material, inexpensive and suitable chemical composition and, moreover, its biodiesel meets the quality parameters in terms of physico-chemical properties. Acknowledgments The authors are grateful to Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP) and Fundação de Amparo à Pesquisa do Estado de Minas Gerais (FAPEMIG) for financial support. Additionally, the authors would like to thank Dr. Cynthia Maria de Campos Prado Manso for revising and editing the text. References [1] Suarez PAZ, Meneghetti SMP. 70th Anniversary of biodiesel in 2007: historical evolution and current situation in Brazil. Quim Nova 2007;30:2068e71. [2] Pousa GPAG, Santos ALF, Suarez PAZ. History and policy of biodiesel in Brazil. Energy Policy 2007;35:5393e8. [3] Rinaldi R, Garcia C, Marciniuk LL, Rossi AV, Schuchardt U. Synthesis of biodiesel: a contextualized experiment proposal for the general chemistry laboratory. Quim Nova 2007;30:1374e80.
[4] Demirbas A. Energy and environmental issues relating to greenhouse gas emissions in Turkey. Energy Convers Manage 2003;44:203e13. [5] Ferrari RA, Oliveira VD, Scabio A. Biodiesel from soybean: characterization and consumption in an energy generator. Quim Nova 2005;28:19e23. [6] Lotero E, Liu YJ, Lopez DE, Suwannakarn K, Bruce DA, Goodwin JG. Synthesis of biodiesel via acid catalysis. Ind Eng Chem Res 2005;44:5353e63. [7] Oliveira JS, Montalvao R, Daher L, Suarez PAZ, Rubim JC. Determination of methyl ester contents in biodiesel blends by FTIR-ATR and FTNIR spectroscopies. Talanta 2006;69:1278e84. [8] Correa SM, Arbilla G. Aromatic hydrocarbons emissions in diesel and biodiesel exhaust. Atmos Environ 2006;40:6821e6. [9] Nguyen NT, Demirel Y. A novel biodiesel and glycerol carbonate production plant. Int J Chem React Eng 2011;9:1e25. [10] Vyas AP, Verma JL, Subrahmanyam N. A review on FAME production processes. Fuel 2010;89:1e9. [11] Abbaszaadeh A, Ghobadian B, Omidkhah MR, Naja&filig G. Current biodiesel production technologies: a comparative review. Energy Convers Manage 2012;63:138e48. [12] Yun H, Wang M, Feng W, Tan T. Process simulation and energy optimization of the enzyme-catalyzed biodiesel production. Energy 2013;54:84e96. [13] Singh SP, Singh Dipti. Biodiesel production through the use of different sources and characterization of oils and their esters as the substitute of diesel: a review. Renew Sustain Energy Rev 2010;14:200e16. [14] Serrano M, Oliveros R, Sánchez M, Moraschini A, Martínez M, Aracil J. Influence of blending vegetable oil methyl esters on biodiesel fuel properties: oxidative stability and cold flow properties. Energy 2013;65:109e15. [15] Crabbe E, Nolasco-Hipolito C, Kobayashi G, Sonomoto K, Ishizaki A. Biodiesel production from crude palm oil and evaluation of butanol extraction and fuel properties. Process Biochem 2001;37:65e71. [16] Demirbas A. Comparison of transesterification methods for production of biodiesel from vegetable oils and fats. Energy Convers Manage 2008;49: 125e30. [17] Berrios M, Gutiérrez MC, Martín MA, Martín A. Application of the factorial design of experiments to biodiesel production from lard. Fuel Process Technol 2009;90:1447e51. [18] Hájek M, Skopal F, Capek L, Cernoch M, Kutálek P. Ethanolysis of rapeseed oil by KOH as homogeneous and as heterogeneous catalyst supported on alumina and CaO. Energy 2012;48:392e7. [19] Kiliç M, Uzun BB, Pütün E, Pütün AE. Optimization of biodiesel production from castor oil using factorial design. Fuel Process Technol 2013;111: 105e8. [20] Veli ckovi c AV, Stamenkovi c OS, Todorovi c ZB, Veljkovi&cacute VB. Application of the full factorial design to optimization of base-catalyzed sunflower oil ethanolysis. Fuel 2013;104:433e42. [21] Mendonça DR, Andrade HMC, Guimarães PRB, Vianna RF, Meneghetti SMP, Pontes LAM, et al. Application of full factorial design and Doehlert matrix for the optimisation of beef tallow methanolysis via homogeneous catalysis. Fuel Process Technol 2011;92:342e8. [22] Rashid U, Rehman HA, Hussain I, Ibrahim M, Haider MS. Muskmelon (Cucumis melo) seed oil: a potential non-food oil source for biodiesel production. Energy 2011;36:5632e9. [23] Allawzi M, Kandah MI. Parametric study of biodiesel production from used soybean oil. Eur J Lipid Sci Technol 2008;110:760e7. [24] Noureddini H, Zhu D. Kinetics of transesterification of soybean oil. JAOCS 1997;74:1457e63. [25] Oliveira D, Luccio MD, Faccio C, Rosa CD, Bender JP, Lipke N, et al. Transesterification of soybean oil and castor oil for biodiesel production. Appl Biochem Biotechnol 2005;121:553e60. [26] Firestone DE. Official methods and recommended practices of The American Oil Chemists Society. Illinois: AOCS Press; 1997. [27] Bondioli P, Della Bella L. An alternative spectrophotometric method for the determination of free glycerol in biodiesel. Eur J Lipid Sci Technol 2005;107: 153e7. [28] Sharma YC, Singh B, Upadhyay SN. Advancements in development and characterization of biodiesel: a review. Fuel 2008;87:2355e73. [29] Lobo IP, Ferreira SLC, da Cruz RS. Biodiesel: quality parameters and analytical methods. Quim Nova 2009;32:1596e608. [30] Santos AGD. Avaliação da estabilidade térmica e oxidativa do biodiesel de algodão, girassol, dendê e sebo bovino. Natal-RN, 183 p. (Tese) e Universidade Federal do Rio Grande do Norte. [31] Candeia RA. Biodiesel de Soja: Síntese, Degradação e Misturas Binárias. João Pessoa- PB, 150 p. (Tese). Universidade Federal da Paraíba; 2008.